The Information's TITV Starlink Takes on Telecom Giants, OpenAI’s Stargate Compute Hurdles, Nvidia Earnings Preview
52 snips
Feb 23, 2026 
Sam Liang, CEO of Otter AI, leads a meeting-intelligence company with 30M users. Gil Luria, tech analyst at DA Davidson, covers Nvidia, chip demand, and AI infrastructure. Anissa Gardizi, cloud reporter, explains OpenAI’s Stargate compute pivot and partner hassles. Theo Waite, SpaceX reporter, breaks down Starlink’s $50 plans, retail push, and telecom challenge. Multiple short conversations on competition, compute, and chips.
AI Snips
Chapters
Transcript
Episode notes
OpenAI Shifted To A Large Oracle Partnership
- OpenAI pivoted from doing all build‑out itself to a hybrid where Oracle would handle a large 4.5 GW commitment.
- OpenAI found raising billions solo hard, so it negotiated a 4.5 GW deal with Oracle to share risk and timelines instead of the initial 10 GW vision.
Compute Limits Shape OpenAI's Product Timeline
- Compute capacity directly constrains OpenAI's product roadmap and training of advanced models.
- Anissa noted delays in settling Stargate likely cost OpenAI time, pushing them to scramble for other cloud capacity from Amazon and Google.
Monitor How Chips Are Financed In Stargate Deals
- Watch how OpenAI structures chip purchase and financing in the Oracle 4.5 GW deal.
- Anissa advised monitoring whether Oracle buys chips, OpenAI supplies them, or creative leasing appears because hundreds of billions in hardware are required.